|
|
Jingyu Song, Paul V Preckel and Michael S Delgado |
|
''Fractional logit estimation under varying spatial resolution'' |
( 2020, Vol. 40 No.4 ) |
|
|
We propose a method for estimating logit regression models in the case that the independent variables are measured at a finer-scale spatial resolution than the dependent variable. Whereas the traditional approach is to aggregate the fine-scale data to the resolution of the dependent variable prior to estimation, we propose integrating the aggregation directly into the regression so as to maximize the value of information contained at the fine-scale resolution. Monte Carlo simulations show reasonable finite sample performance and that the traditional approach is biased. Our estimator is applicable in many cases that use remotely sensed or GIS data, such as land use problems. |
|
|
Keywords: Logit regression, spatial resolution, grid cells |
JEL: C1 - Econometric and Statistical Methods: General Q0 - Agricultural and Natural Resource Economics; Environmental and Ecological Economics: General |
|
Manuscript Received : Jun 23 2020 | | Manuscript Accepted : Nov 14 2020 |
|